Method and system for automatically producing video highlights
Methods and systems are provided for automatically producing highlights videos from one or more video streams of a playing field. The video streams are captured from at least one camera, calibrated and raw inputs are obtained from audio, calibrated videos and actual event time. Features are then extracted from the calibrated raw inputs, segments are created, specific events are identified and highlights are determined and the highlights are outputted for consumption, considering diverse types of packages. Types of packages may be based on user preference. The calibrated video streams may be received and processed in real time, periodically.
1. A method for automatically producing highlight videos from one or more video streams of a playing field, comprising:
receiving calibrated raw inputs captured by at least one camera positioned proximate to the playing field, wherein the calibrated raw inputs include audio, calibrated video and actual event time and wherein the audio and the calibrated video are synchronized;
extracting features to transform the calibrated raw inputs into feature vectors by applying low-level processing, wherein said low-level processing comprises utilizing pre-existing knowledge regarding points within the field of view of said at least one camera and therefrom identifying and extracting features, the pre-existing knowledge including: knowledge about areas of the playing field, knowledge about certain players, and knowledge about how various players move around the playing field;
creating segments from the feature vectors and identifying specific events in each one of the segments by applying rough segmentation;
determining whether each one of the events is a highlight by applying analytics algorithms; and,
outputting the highlights for consumption and determining types of packages to be outputted for consumption.
2. The method of claim 1 , wherein the receiving of calibrated raw inputs is performed in real-time every predefined time period.
3. The method of claim 1 , wherein the types of packages include at least one of:
condensed game;
replays;
a highlight film; and,
a personal clip.
4. The method of claim 1 , further comprising determining the types of packages based on user preferences.
5. The method of claim 1 , wherein determining whether each of the events is a highlight is further based on deep learning process.
6. The method of claim 1 , wherein the audio includes ambient audio.
7. The method of claim 1 , wherein the audio includes commentator audio.
8. The method of claim 1 , wherein the captured raw inputs include at least one of:
specific players;
various areas of the playing field;
scoreboard;
game clock;
motion pictures of players' activities in the playing field;
relative times; and,
user collected statistics.
9. The method of claim 1 , further comprising storing the raw input in a storage medium, and providing access though a communication network.
10. A system for automatically producing highlight videos from one or more video streams of a playing field, comprising:
a home server including one or more processors;
one or more non-transitory computer-readable media storing one or more sets of instructions which, when executed by one or more processors, cause the one or more processors to perform:
receiving calibrated raw inputs captured by at least one camera positioned proximate to the playing field, wherein the raw inputs include audio, calibrated video and actual event time and wherein the audio and the calibrated video are synchronized;
extracting features to transform the calibrated raw inputs into feature vectors by applying low-level processing, wherein said low-level processing comprises utilizing pre-existing knowledge regarding points within the field of view of said at least one camera and therefrom identifying and extracting features, the pre-existing knowledge including: knowledge about areas of the playing field, knowledge about certain players, and knowledge about how various players move around the playing field;
creating segments from the feature vectors and identifying specific events in each one of the segments by applying rough segmentation;
determining whether each one of the events is a highlight by applying analytics algorithms; and,
outputting the highlights for consumption and determining types of packages to be outputted for consumption.
11. The system of claim 10 , wherein the receiving of the calibrated raw inputs is performed in real-time every predefined time period.
12. The system of claim 10 , wherein the types of packages include at least one of:
condensed game;
replays;
a highlight film; and
a personal clip.
13. The system of claim 10 , further comprising determining the types of packages based on user preferences.
14. The system of claim 10 , wherein determining whether each of the events is a highlight is further based on deep learning process.
15. The system of claim 10 , wherein the audio includes ambient audio.
16. The system of claim 10 , wherein the audio includes commentator audio.
17. The system of claim 10 , wherein the captured raw inputs include at least one of:
specific players;
various areas of the playing field;
scoreboard;
game clock;
motion pictures of players' activities in the playing field;
relative times; and,
user collected statistics.
18. The system of claim 10 , further comprising:
a network communication; and,
wherein the one or more processors further perform:
storing the calibrated raw inputs in a storage medium; and,
providing access though the communication network.